Slightly tangent: Indian government has mandated all ISPs to upgrade all customers as IPv6 compatible by 31st December 2022. So India will be 100% IPv6 compatible by end of the year.
Once that happens, I wouldn't be surprised if IPv4 is deprecated along with 2G.
> For the past 10 years [US] T-Mobile has worked towards creating an IPv6 environment and we are now getting very close to our goal. Stephan presents learning on how to successfully enable IPv6-only using DNS64 with or without 464XLAT. He will do a live demo of the different IP interfaces on an Android handset. Finally, he will discuss and give some best practices on how to handle DNS, applications, and websites that are having issues with DNS64.
iPhones on T-Mobile do not use 464XLAT, they use native IPv6 only. Some system frameworks like URLSession can do the NAT64 translation of hard-coded IPv4 addresses but it still goes over IPv6. In the video linked above: https://youtu.be/nNMNglk_CvE?t=1651
Precisely what I meant is that unless you use very low level networking functions in your app, you will have no problems with that requirement even if your server is ipv4-only, as is my case.
Maybe I say something stupid, but could a "IPv4+/IPv6 ready" protocol maybe ease the transition?
What I mean is the following: as IPv4 is 32bits and IPv6 is 128bits, you could write a subset of IPv6 in the format nnn.nnn.nnn.nnn where n is a digit and you would allow nnn to have values from 0-999 instead of 0-255 as it is today for IPv4.
I will explain one nnn part, the others work the same.
0~255 means 0~255.0.0.0
256~511 means 255.0~255.0.0
512~767 means 255.255.0~255.0
768~999 means 255.255.255.0~234
You have 4 of these nnn's, so 4 x 4 bytes = ( a little less than ) 16 bytes = 128 bits, with the restriction that not all combinations can be made.
But if current hardware could get patched firmware to accept/use these kind of "IPv6 ready" addresses, then the IP address could stay the same (or at least could easily be converted to a IPv6 address) if a true conversion to IPv6 only network would take place. I can imagine that this would give possibilities for a smoother transition.
This is the one market where I would be happy for the cryptobros to move in and start hoarding. I’d almost forgive them for NFTs if IPv4-hodling turned out to be the thing that finally catalyzes the move to IPv6.
Unfortunately you cannot simply HODL IPv4 because authorities can come over and take you blocks back if they are not used.
I can see an incentive to form a small time ISP as an investment though but that is also quite a hands on investment unlike NFT or some coins where you buy and keep.
> Unfortunately you cannot simply HODL IPv4 because authorities can come over and take you blocks back if they are not used.
No longer true, depending on the RIR and their exact policies. At least RIPE and ARIN see leasing blocks as proper usage. The only thing you might run into a snag with would be if you had zero customers leasing blocks within the region the RIR is responsible for. For example, if you have a /19 assigned to ARIN right now and want to lease it to a European entity announcing it from Europe, you might need to transfer it to RIPE so the proper party gets paid membership dues.
It's certainly more hands on than holding equities (or crypto), much like renting out a few apartments is more effort than simply holding an REIT. Quite doable for anyone with an ISP/infrastructure background and capital to put into acquiring ipv4 blocks. It's actually much easier then renting out real estate (I do both), as it's mostly just jumping on tickets that need you to click a couple buttons every so often. If you find some stable/large lessees, it's more or less set it and forget it for N number of years.
I'm not sure I'd advise people to jump into this market when IPv4 is hitting $50 per address, but I've also thought the "top" was in years ago so what do I know :)
I've pondered for a few years now somehow offering "IPv4 investors" fractional ownership of an LLC stood up to lease out address blocks - I just never had the time to get through the legal/tax complexity of having 50-100 partners all pooling capital. Banks are surprisingly willing to finance these acquisitions as well, assuming you have an industry track record.
> I'm not sure I'd advise people to jump into this market when IPv4 is hitting $50 per address, but I've also thought the "top" was in years ago so what do I know :)
You and me both. I sold half my IPv4 holding at $20 (having bought in at $8), and am both annoyed and glad that I only did half, since I can now afford to sell the remaining 25% at a high premium, and maintain my remaining /24 and ASN as a vanity project for many decades.
> Unfortunately you cannot simply HODL IPv4 because authorities can come over and take you blocks back if they are not used.
Define "not used". I know several LIRs that have in excess of /20s allocated to them, and do not advertise their ranges to the (Internet) DFZ, but instead use them on private interlinks between themselves and other third parties.
The only requirement to be given address space by an RIR is a requirement for global uniqueness.
Just because you don't see it on the Internet (or any other network you use), doesn't mean it's not in use.
IPv4 is knocking on death's door, and the value will drop once the DoD gives (6 years) notice that they've migrated all their services to IPv6, thus making 10? /8s available for sale and:
1. flooding the market
2. proving that there's nothing left holding back to run an ipv6-only network, resulting in wide scale migrations for networks that haven't migrated yet.
I'm not speaking for ARIN but their CEO John Curran told me ARIN has never revoked address resources for any reason other than non-payment or fraud. As long as your original application for the resources was made in good faith and you continued to pay the registration fees, they would not take them away even if you are not using them anymore. This seems logical for functional IPv4 markets where the non-use/non-need is explicit in any resource being offered for sale. The registered organization does still need to be "in business" (registered in good standing with the state corporation commission or equivalent). If the entity is dissolved then the resources automatically revert back to ARIN and it may be difficult or impossible to sell to a third party.
It is important to keep ARIN updated with any organization name changes, reorganization, or M&A actions as they happen. Going back 10+ years to update those changes, especially if there are several, can be extremely difficult. In large companies the network folks that manage ARIN resources are often not even be aware of those changes until they interact with ARIN again. That has bitten me several times in my career.
In the old days before IPv4 runout they could and would refuse to issue additional IP blocks if you did not show sufficient utilization of your most recent allocation. It was also not unusual for them to reduce the allocation size requested based on how fast you had utilized your most recent block.
Hoarding IP addresses probably is fraud because to get IPs you have to swear that they will be put into use within a year (IIRC). I don't consider IP leasing to be hoarding because the IPs are in use.
What counts as “not used”? If I set up some minimal websites hosting one ape picture per IP address, is that used? Do the registries have a concept of frivolous use?
I’ve got a /24 that hasn’t had any use at all since about 1998, and in fact is registered to a company that no longer exists, with me as the only point of contact. Nobody has ever come for it. I wouldn’t be overly upset if they did, as I don’t have any use for it, but I’ve never felt like proactively giving it up either.
But even setting one ape per IP address is still quite hands on and involved investment, you have to know how to do it or you have to hire and pay someone who knows how to do it.
Then you still have some operation cost as you probably cannot assign whole IP block to a single server and keeping server running still costs money.
So I was only think about situation where you simply buy like /24 block and keep it.
This is a great article (as expected considering the author). It is a much more nuanced take on the whole IPv4 vs IPv6 topic than the recurring heated discussions on HN.
> We continue to apply the economics of abundance to IPv6 address markets.
I'm an ordinary retail customer of my ISP, which is IP6-enabled. So they allocated me a /56 prefix.
That certainly looks to me like "the economics of abundance". Even if every device, appliance, light-fitting and even door in my home had its own global IP address, I'd never use more than a tiny part of that allocation.
Still, it's the "economics of abundance" to automatically dole out huge amounts of address space to someone who can't demonstrate the need for more than a few dozen addresses.
A /56 is a good fit for a residential ISP. Initially some ISPs gave out a /48.
A /56 gives you 256 subnets. A /60 only 16. I'm happy they are erring on the side of potential innovation. ISPs have no idea how IPv6 will be used.
A few years back, I was running a ton of various services from home, various APs, community mesh network, services over VPN. It worked well on a /56. (Thank you Teksavvy)
If You're running multiple APs (with IP6 over wifi) and a community mesh, you have a demonstrable need for space. I have a single AP and a few strands of CAT5.
Can you even run IP6 over wifi? I've never looked into it. It's never crossed my mind to try to use any public IP address on a wifi network.
The wi-fi access point doesn't even need to be IPv6-aware, I'm pretty sure. Most can be set to a "bridge" mode that should work fine (note that I did not test this personally).
But even cheap wireless routers I bought 10 years ago support IPv6, and I doubt you really want to use that on a modern Wi-Fi network.
Once that happens, I wouldn't be surprised if IPv4 is deprecated along with 2G.